Ocean Drilling Program: Science Operator - 69. …...In September 1989, an ocean broadband digital downhole seismometer (OBDS) was installed at Hole 794D in the northern Yamato Basin

Tamaki, K., Suyehiro, K., Allan, J., McWilliams, M., et al., 1992Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 127/128, Pt. 2

69. DETAILED CRUSTAL STRUCTURE OF NORTHERN YAMATO BASIN1

M. Shinohara,2,3 N. Hirata,2 H. Nambu,2 K. Suyehiro,4 T. Kanazawa,5 and H. Kinoshita6

ABSTRACT

In September 1989, an ocean broadband digital downhole seismometer (OBDS) was installed at Hole 794D in the northernYamato Basin by the Ocean Drilling Program (ODP). We conducted a high-resolution refraction/reflection experiment using theOBDS and nine ocean-bottom seismometers (OBS' s), which were deployed by the Tansei-maru. We continuously recorded signalsfrom the OBDS in digital form on the JOIDES Resolution for about 2 days.

The OBDS had a three-component feed-back-type accelerometer. The signals were sampled at 80 Hz with 16-bit resolution.The OBS's recorded signals for about 2 weeks by the direct analog recording method. The five OBS's, which were closely spacedevery 2800 m near Hole 794D, and the OBDS formed a three-dimensional sub-array. The air-gun profiles consisted of two circlesand two linear lines. Air guns were shot every 60 s at 10 MPa with two 9-L or a single 17-L chamber. Signals from a single channelhydrophone streamer were also recorded with a personal computer system in digital form.

We determined the velocity structure near the profiles using the OBDS and four OBS's that were deployed every 25 km. Atwo-dimensional (2-D) velocity model was constructed as follows: first, one-dimensional velocity structures under each OBSwere derived by the tau-/> inversion method; second, 2-D velocity structures under the linear profiles were derived by the forward2-D ray tracing method.

A sedimentary layer with a small velocity gradient exists just beneath the seafloor. The layer has a velocity of 1.7 km/s andis 200-700 m thick. The layer corresponds to the transparent layer on the reflection profile. An acoustic basement with a P-wavevelocity of 4.2^.5 km/s underlies the sedimentary layer. The acoustic basement layer has a velocity gradient of 0.25 s~'and is2-3 km thick. A layer with a velocity of 6.2 km/s underlies the acoustic basement layer. The velocity gradient in the 6.2 km/slayer is less than 0.2 s"1. We could not determine the thickness of the 6.2 km/s layer because of the lack of deep penetration ofrefracted waves down to the Moho. However, the thickness of the 6.2 km/s layer is estimated to be more than 6 km based on thedepth of penetration of refracted wave through the 6.2 km/s layer.

The upper crustal structure is similar to that found in the southern Yamato Basin, where the entire crust is neither typicaloceanic nor continental.

INTRODUCTION

The Japan Sea is one of the backarc basins in the western Pacific.It comprises the Japan Basin in the northwest and the Yamato Basinin the southeast, and is divided by the Yamato Rise (Fig. 1). Theseafloor depths of the two basins are nearly constant at about 3500 mand 2500 m, respectively.

A number of models have been proposed to explain the formationand evolution of the Japan Sea from the seismic structure of thelithosphere (Abe and Kanamori, 1970; Evans et al., 1978), magneticanomalies (Isezaki et al., 1971; Isezaki and Uyeda, 1973; Isezaki,1975; Seama and Isezaki, 1990), paleomagnetic evidence on land(Otofuji and Matsuda, 1984; Otofuji et al. 1985; Otofuji and Matsuda,1987), and heat-flow data (Yasui et al., 1968; Yoshii, 1972; Tamaki,1986; Yamano et al., 1987). One of the main objectives of Legs127/128 drilling was to obtain ground-truth evidence to solve thismajor problem.

Any valid model should be able to explain the seismic structureof the Japan Sea crust, which appears to be highly variable. Theseismic structure of the Yamato Rise suggests a continental crustalstructure, though its thickness is undetermined (Murauchi, 1972;Ludwig et al., 1975). The Japan Basin has a typical oceanic crustalthickness of 8 km (Murauchi, 1972; Ludwig et al., 1975). The seismiccrustal structure of the Yamato Basin is 14 km thick, which is twice

Tamaki, K., Suyehiro, K., Allan, J., McWilliams, M., et al., 1992. Proc. ODP, Sci.Results, 127/128, Pt. 2: College Station, TX (Ocean Drilling Program).

2 Department of Earth Sciences, Faculty of Science, Chiba University, Chiba 260,Japan.

Present address: Ocean Research Institute, the University of Tokyo, Tokyo 164,Japan.

4 Ocean Research Institute, the University of Tokyo, Tokyo 164, Japan.5 Faculty of Science, the University of Tokyo, Tokyo 113, Japan.6 Earthquake Research Institute, the University of Tokyo, Tokyo 113, Japan.

as thick as a typical oceanic basin. However, the velocity in each layeris similar to that of oceanic crust except that the P-wave velocity inthe uppermost mantle (Pn velocity) is lower than a normal oceanic Pn

velocity (Murauchi, 1972; Ludwig et al., 1975; Hirata et al., 1987;Katao, 1988; Chung et al., 1990).

An anisotropic velocity model for the uppermost mantle beneaththe Japan Sea has been proposed by Okada et al. (1978). The anisot-ropy in the uppermost mantle is confirmed by an OBS array study inthe southern Yamato Basin, where the maximum velocity direction isperpendicular to the strike of the northwest Japan arc (Hirata et al.,1989). The crustal seismic anisotropy in the studied region is reportedby Hirata et al. (this volume).

The use of OBS's in refraction experiments has been shown toimprove the quality and resolution of the seismic structural model asevidenced in the 1985 OBS and multichannel seismic profiling(MCS) experiment in the southwestern Yamato Basin (Hirata et al.,1987, 1989). The use of a downhole seismometer in recent refractionexperiments has proven to be useful in obtaining higher quality datasets and a high-resolution seismic structure (Stephen et al., 1983;Swift et al, 1988; Little and Stephen, 1985; Stephen and Harding,1983; Jacobson et al., 1984; Shearer et al., 1986a; Whitmarsh et al.,1986; Duennebier et al., 1987; Byrne et al., 1987; Bee and Bibee,1989), and in studying of seismic anisotropy and lateral heterogeneity(Stephen, 1985; 1988; Swift and Stephen, 1989; Shearer and Orcutt,1985; Shearer et al., 1986b). A refraction experiment using both adownhole seismometer and OBS's, which promises greater resolutionof seismic structures especially in laterally heterogeneous region, canthus contribute important details relevant to understanding the tec-tonic evolution of a particular region.

In September 1989, an ocean broadband downhole seismometer(OBDS) was deployed at Hole 794D in the northern Yamato Basin ofthe Japan Sea by the JOIDES Resolution during Ocean DrillingProgram (ODP) Leg 128 (Ingle, Suyehiro, von Breymann, et al.,1990). At this time, nine OBS's were also deployed around Hole 794D

1075

M. SHINOHARA ET AL.

45° N

40c

300 km

{V Yamato Ri

K θ r e a l 1.20001000

130°E 135C 140c

Figure 1. Map showing locations of drill holes in the Japan Sea (Ingle, Suyehiro, von Breymann, et al., 1990).Contour interval is 1,000 m. Double circles denote holes drilled by Leg 128, ODP (Sites 794,798 ,799). The OBDSwas installed at Hole 794D. Circles and squares denote holes drilled by Leg 127, ODP, and Leg 31, DSDP,respectively. PI and P4 are two ship refraction experiment lines reported by Murauchi (1972). YB7 shows locationof an OBS deployed in 1985 reflection and refraction seismic experiment (Hirata et al. 1987).

(Fig. 1). A controlled source seismic refraction/reflection experimentusing air guns , the OBDS, OBS's, and a hydrophone streamer wascarried out with the support of the Tαnsei-mαru, Ocean ResearchInstitute (ORI), the University of Tokyo, to obtain a high-resolutionseismic crustal structure in the northern Yamato Basin. In the presentpaper we report the detailed analysis of the crustal structure nearSite 794 and focus on vertical and lateral heterogeneity of the crust.

EXPERIMENTAL CONFIGURATION

Observation System

OBDS System

The OBDS was installed on September 26, 1989. Hole 794D is734 m below the seafloor (mbsf) at a water depth that is 2807 m belowsea surface. A sedimentary section of Hole 794D was cased through560 mbsf. The seismometer was clamped at 715 mbsf in a hardigneous rock section. Real-time recording of digital data from thethree-component feed-back-type accelerometer was performedaboard the JOIDES Resolution with time calibrated against JapanStandard Time by the radio signal JJ Y. A detailed description of theOBDS system is reported by Suyehiro et al. (this volume).

OBS System

Two types of OBS's, both of which are free-fall and transpondertriggered pop-up types, were used in this study. One type of OBS wasdeveloped at the Geophysical Institute of the University of Tokyo, in

cooperation with the Laboratory for Ocean Bottom Seismology,Hokkaido University (Yamada, 1980; Kanazawa, 1986). The othertype was developed at Department of Earth Sciences, Chiba Univer-sity, from the University of Tokyo type OBS (Matsuda et al., 1986),and contains a hydrophone channel.

Three-component velocity sensors with a natural frequency of4.5 Hz are held by a gimbal mechanism. Signals are recorded on audiocassette tape by the direct analog recording (DAR) method. Becauseof the stability of the recording mechanism and a small temperaturevariation at the deep seafloor, the accuracy of timing in recordingthroughout the observation period is estimated to be less than 0.05 s.All the analog data of OBS's were digitally processed on shore afterthe cruise (Urabe and Hirata, 1984).

Digital Single-channel Reflection Experiment System

Air-gun signals were also recorded by a digital single-channelhydrophone streamer towed by the Tansei-maru (Shinohara et al.,1989). The structure of the sedimentary layers was determined pre-cisely from reflection records with a large dynamic range. As theacoustic basement is rugged, the high dynamic range also serves toenhance the resolution of the deeper structure.

Positioning

Ship positioning of the Tansei-maru, which acted as the shootingvessel, was accomplished using the Global Positioning System (GPS)and Loran-C. A detailed discussion of navigation accuracy can be

1076

YAMATO BASIN CRUSTAL STRUCTURE

found in Hirata et al. (this volume). Here, we describe the outline. Thepositioning data from the navigation system were stored on a micro-computer every 4-15 s. Scatter in position data indicates that theaccuracies of the GPS and the Loran-C are about 20 m and 200 m,respectively. First, we determined the positions where the air gunswere shot, by interpolation of GPS data, where available. The posi-tions determined by Loran-C data were found to be a linear translationof GPS data. Hence, when GPS data were not available, the positionsdetermined by Loran-C were used with appropriate corrections. Theaccuracy of all the positions is estimated to be about 20 m.

Because the OBS's are a free-fall type, positions of the OBS's atthe deep seafloor may differ from their launching positions. There-fore, the positions of the OBS on the seafloor were relocated usingthe traveltimes of direct water waves within 10 km of the OBS.Typically, 200 of direct-wave traveltimes were used for each OBSrelocation. The estimated accuracy of the instrument location is about150 m, which is comparable to a wavelength of the direct water wave.

Real-time Experiment (RTE)

Nine OBS's were deployed to obtain seismic records from the airguns (Fig. 2). One OBS launched near the OBDS recorded digital datafor 9 hr, and the others recorded analog data for about 2 weeks. Allthe OBS's were successfully retrieved after the RTE. Regrettably, thedigital-type OBS suffered from hardware trouble and recorded noseismic data. A three-dimensional (3-D) sub-array was formed by fiveclosely spaced (2800 m) OBS's and the OBDS. The locations and theperiods of observation for all OBS's are listed in Table 1.

The air-gun shooting profiles for the controlled source seismicexperiment consisted of two concentric circles and two cross lines.Radii of the two circle profiles are 9 km (Inner Circle) and 18 km(Outer Circle). The east-west trending line (Line EW) is 90 km long,the other north-south line (Line NS) perpendicular to Line EW is70 km long. The center of the circles and also the cross point of thetwo lines is the position of the OBDS. Air guns were shot every 60 sat 10 MPa by either two 9-L or a single 17-L air gun (Table 2). Thespacing between successive air gun shots was about 150 m. Duringthe 48-hr time period of shooting, data were recorded continuouslyby the OBDS. In this paper, we use records from four OBS's (JRT1,4, 5, 8) and the OBDS to determine the seismic P-wave velocitystructure along the two linear lines (Line EW, NS).

DATA

Records of Digital Single-Channel (DSC) ReflectionSystem

The single-channel digital reflection records were band-pass fil-tered (30-100 Hz) and each trace shown was normalized by itsmaximum amplitude to enhance reflection signals.

Line EW (Fig. 3)

The seafloor is deeper around the position of the OBDS, and showsmorphologically high relief at the positions of JRT1 and JRT5. Thethickness of the sedimentary layer at the OBDS is about 0.8 s intwo-way traveltime. The acoustic basement shows morphologicallyhigh relief resembling a knoll at 13 km to the east and 22 km to thewest of the OBDS.

Line NS (Fig. 4)

The seafloor deepens northward from 15 km south of the OBDS,where morphologically high relief is seen. The thickness of thesedimentary layer is almost constant in the northward section of theOBDS. The acoustic basement also shows morphologically highrelief 15 km south of the OBDS.

40°30'N

Northern Yamato BasinOBDS-OBS array

40°0•

39°50'

137°40'E 138°0' 138°50'

30km

Figure 2. Local bathymetry surrounding ODP Hole 794D and locations of theOBDS and nine OBS's. Contour interval is 100 m. Five OBS's spaced closelyevery 2800 m, and the OBDS formed a 3-D sub-array. The digital-type OBSwas deployed at the same position as JRT4. Air-gun profiles (solid line) consistof two concentric circles and two perpendicular lines.

Records of the Seismometers

We will discuss seismograms recorded on the high-gain, vertical-component geophone. The band-pass frequency is from 5 to 15 Hz.Each trace shown is normalized by its maximum amplitude. Thedistance is measured from the position of the OBDS which is nearlyat the center of the profiles. All of the seismometer records for LineNS have generally higher signal-to-noise ratios than the records forLine EW due to better weather conditions.

Line EW

OBDS (Fig. 5)

Between the ranges of 2 and 6 km east of the OBDS, the first-ar-rival phase has an apparent velocity of 4.3 km/s. At offset distancesgreater than 30 km, the apparent velocity of the first arrivals is7.2 km/s. The small amplitudes of the first arrivals at 10 km areattributed to a knoll. We have no data from 6 to 10 km because of atechnical problem. In the western section, the apparent velocity of thefirst arrivals is 3.8 km/s at offset distances from 1 to 5 km; 5.0 km/sat offset distances from 5 to 13 km; and 7.1 km/s at offset distancesgreater than 13 km. A triplication is seen at a distance of 10 km. Thefirst arrivals are seen up to 40 km west of the OBDS.

JRT4 (Fig. 6)

JRT4 is the OBS just above the OBDS. In the east section, theapparent velocity for the first arrivals is 4.3 km/s at offset distancesless than 7 km and 6.4 km/s at offset distances greater than 7 km. Inthe west section, an apparent velocity for the first arrivals is 4.7 km/sat distances less than 12 km and 6.3 km/s at offset distances greaterthan 12 km. The first arrivals are seen up to 27 km east of the OBSand 25 km west of the OBS. The small amplitude of the first arrivalsat distances from 9 to 13 km east of JRT4 result from a knoll.

JRT1 (Fig. 7)

JRT1 was located at about 25 km east of the OBDS. On the westernside of JRT1, the apparent velocity of the first arrivals is 4.4 km/s atoffset distances less than 12 km and 5.9 km/s at offset distances greater

1077

M. SHINOHARA ET AL.

Table 1. Locations of OBS's and the OBDS and period of observation(Japan standard time).

OBSCode

Latitude (N))eg. Min.

Observation period

Longitude (E) Depth Deployment RetrievalDeg. Min. (m) Sept. 1989 Sept. 1989

JRTIJRT2JRT3JRT4

aJRT4HJRT5JRT6JRT7JRT8OBDS

40404040

4040404040

18.0713.0412.3411.37

h

4.8812.6314.0222.9411.17

138 28.72138 17.24

138 15.75

138 14.03

137138138138

138

59.5912.6311.915.5914.13

2419284128202881

24402908293132133533

24, 23:5225,01:0325,01:2925,02:1125,02:1425, 06:3325, 02:4925,03:1925, 04:37

30, 15:1130, 13:2030, 11:0930, 09:56

29, 16:5530,09:1630, 07:3029, 20:07

aA digital recording OBS.No positioning due to no signal record.

cNo time record of retrieval.

Table 2. Shooting periods and total number of shots in the refraction/reflectionexperiment (Japan standard time).

Line Time start shootingSept. 1989

Time end shootingSept. 1989

Total number of shots

Inner circleOuter circle

Line-EWLine-NS

27, 07:36:0527, 15:13:0528, 06:00:0528,23:17:55

27, 13:52:0528, 03:45:0528, 15:30:0529, 05:35:55

377596571379

than 12 km. A triplication is seen at an offset distance of 10 km. Thefirst arrivals are seen up to 40 km to the east of JRTI. On the easternside of JRTI, the apparent velocity of the first arrivals is 5.0 km/s outto the end of the profile at 19 km.

JRT5 (Fig. 8)

JRT5 was located at about 25 km west of the OBDS. On the easternside of JRT5, the apparent velocity of the first arrivals is 3.6 km/s atoffset distances less than 9 km and 6.4 km/s at offset distances greaterthan 9 km. A triplication is also seen at an offset distance of 10 km.The first arrivals are seen up to 25 km to the west of JRT5. On thewestern side of JRT5, the apparent velocity of the first arrivals is3.7 km/s at offset distances less than 11 km and 5.5 km/s out to theend of the profile at 19 km.

Line NS

OBDS (Fig. 9)

In the southern section, the apparent velocity of the first arrivals is5.0 km/s at offset distances less than 9 km and 5.8 km/s at offset distancesgreater than 8 km. The first arrivals are seen out to the end of the profile(37 km). Small amplitudes for the first arrivals are seen at offset distancesfrom 8 to 17 km due to a knoll. In the northern section, the apparentvelocity of the first arrivals is 3.4 km/s at offset distances less than 10 kmand 7.5 km/s at offset distances greater than 10 km. A triplication isclearly seen at an offset distance of 10 km. Refracted arrivals from thecrust are seen out to the end of the profile (36 km).

JRT4 (Fig. 10)

In the southern section, the apparent velocity of the first arrivalsis 5.0 km/s at offset distances from 4 to 10 km and 6.3 km/s at offsetdistances greater than 18 km. Small amplitudes for the first arrivalsdue to a knoll are also seen at offset distances from 10 to 18 km. Thefirst arrivals are seen at up to an offset distance of 30 km. In the

northern section, the apparent velocity of the first arrivals is 4.3 km/sat offset distances less than 13 km and 6.7 km/s at offset distancesgreater than 13 km. A triplication is also seen at an offset distance of10 km. The first arrivals are seen out to an offset distance of 30 km.

JRT8 (Fig. 11)

JRT8 was located at 25 km south of the OBDS. In the southernside of JRT8, the apparent velocity of the first arrivals is 4.1 km/s atshort offset distances, 5.5 km/s at offset distances from 9 to 14 km,and 7.2 km/s at offset distances greater than 14 km. First arrivals areseen out to an offset distance of 35 km. A triplication is also seen atan offset distance of 10 km. A later phase that seems to be the reflectedarrival from the lower crust with an apparent velocity of about 8.0km/s, is seen at offset distances greater than 45 km. On the northernside of JRT8, the apparent velocity of the first arrivals is 4.8 km/s.

Comparison Between the OBS Records andthe OBDS Records

The differences between the records of the OBS's and those of theOBDS are as follows:

1. The data recorded by the OBDS have better signal-to-noiseratios than those recorded by the OBS. Therefore, refracted arrivalson the OBDS can be seen out to greater distances than on the OBS.

2. The apparent velocity of the direct wave recorded by the OBDSis about 3.0 km/s far from the OBDS. The direct wave is defined asthe wave that propagates through the media above the receiver, whichin this case includes the sediment layer.

3. Because the velocity gradient of the sedimentary layer is largerthan that of water, the range in which the OBDS records the directwaves is narrow.

4. There are no coda waves after the direct-wave arrivals in therecords of the OBDS.

5. Converted S-waves are clearly seen in the records of the OBDS(Fig. 12). Comparison of the horizontal component records revealsthat P-wave arrivals are also clearly seen on the OBDS. These resultsare indicative of a strong mechanical interaction (coupling) betweenthe OBDS and the ground, as compared to that between an OBS andthe ground.

ANALYSES AND RESULTS

One-dimensional Shallow Velocity Structure

We derived one-dimensional (1-D) velocity structures beneatheach OBS from the individual records. First, the seismic wave fieldin x-t (distance-traveltime) domain on one side of the OBS, is trans-formed into the tau-p (intercept time-ray parameter) domain (Dieboldand Stoffa, 1981; Stoffa et al., 1981). To map seismic data in x-tdomain to the domain of tau-p, we performed a slowness stack or rayparameter stack for linear trajectories through the, x-t data. Semblancewas used instead of a slowness stack to suppress the spatial aliasingand to get good signal-to-noise ratios in tau-p domain. Semblance hasbeen used in detecting coherent arrivals across an array (Neidell andTaner, 1971). After seismic data were mapped to tau-p domain,trajectories of refractions and wide-angle reflections are picked fromthe records. A P-wave velocity-depth function is then derived fromthe trajectory by the tau-sum inversion method (Diebold and Stoffa,1981). Hereafter, we call this procedure the tau-p method.

We used the low-gain vertical geophone component at distancesfrom 0 to 10 km to derive the shallow structure and the high-gainvertical component at distances from 10 to 20 km to derive the deeperstructure. In the tau-p domain, this corresponds to trajectories wherep is greater than 0.2 s/km on the low-gain component, and trajecto-ries where p less than 0.2 s/km on the high-gain component. Thepicking errors of the trajectories are about 0.1 s in tau domain. This

1078


wJRT5

OBDSJRT4 JRT1

I 4-CD

1>N

?§

10 20 30 40 50 60 70 80

Distance (km)

Figure 3. Digital single-channel seismic reflection record for Line EW obtained simultaneously with the refraction experiment. Distance along the profile is positiveeastward. Band-pass filtering is 30-100 Hz. Each trace is normalized by its maximum amplitude. The positions of the OBDS and the OBS's (JRTl, JRT4, JRT5)are marked.

JRT8OBDSJRT4

CΛ

0 30 40

Distance (km)

50 60 70

Figure 4. Digital single-channel seismic reflection record for Line NS. Distance is positive southward. The positions of the OBDS and the OBS's (JRT4, JRT8)are marked. See Figure 3 for detailed explanation.

1079

M. SHINOHARA ET AL.

OBDS

WMammBSBm

-40 -30 -20 -10 0 10

Distance (km)

40

Figure 5. Record section for the vertical component, high-gain channel of the OBDS for Line EW. Reduction velocity is 5.0 km/s. Band-pass filtering is 5-15 Hz.Each trace is normalized by its maximum amplitude. First arrivals are seen at greater distances on the OBDS than on the OBS's and no coda waves are observedafter the direct waves on the OBDS.

JRT4

04--40 -30 -20 -10 0 10

Distance (km)

Figure 6. Record section of OBS JRT4 for Line EW. See Figure 5 for detailed explanation.

20 30

value of picking error is considered to correspond to a few hundred metersof depth.

The structure derived by the tau-p method assumes horizontallyhomogeneous layers. We applied the tau-p method to records on bothsides of the OBS, thus resulting in two velocity-depth functions foreach OBS profile. Differences in the two velocity-depth functions foreach side were taken to represent lateral heterogeneity in the crust.

The 1-D velocity-depth functions were derived from records offour OBS's: JRT1, JRT4, JRT5, and JRT8 (Figs. 13 through 17). Asan example, the velocity-depth functions are compared with theassociated single-channel reflection profiling records in Figure 18.The velocity in the uppermost layer just below all OBS's is 1.7 km/s

with a small velocity gradient with depth. The 1.7 km/s layer is about700 m thick under JRT4, about 500 m under JRT1, about 300 m underJRT5, and 700 m under JRT8. The 1.7-km/s layer corresponds to theacoustic transparent sedimentary layer seen in the reflection records.

A layer with a P-wave velocity of about 4.5 km/s underlies thesedimentary layer under JRT1, JRT4, and JRT8. The vertical gradientof velocity in the 4.5-km/s layer is about 0.25 s"1. The velocity of theuppermost part of the 4.5-km/s layer increases rapidly from 2.0 to4.5 km/s within the uppermost 300 m. In the reflection records, therapid increase in velocity corresponds to a clear reflector of theacoustic basement. The increase in velocity under JRT5, however, isdifferent from others. The velocity just below the sedimentary layer

1080


JRT1W

-30 -20 -10 0 10

Distance (km)

Figure 7. Record section of OBS JRTl for Line EW. See Figure 5 for detailed explanation.

20 30 40

JRT5

0-40 -30 -20 -10 0 10

Distance (km)

Figure 8. Record section of OBS JRT5 for Line EW. See Figure 5 for detailed explanation.

20 30

is 2.0 km/s with a gradient of 5.0 s '. The 4.5-km/s layer underlies the2.0-km/s layer and is 1500 mbsf. The velocities and the velocitygradients in the 4.5-km/s layer are similar for all the OBS's.

Two-dimensional Velocity Structure

Two-dimensional (2-D) velocity structures along lines EW and NSwere derived by 2-D ray tracing using first arrivals and later phasesto derive shapes of velocity interfaces and velocities at deeper levels.The first refracted arrivals for all OBS's were picked from the records(Fig. 19). The apparent velocities of the first arrivals of all the OBS'sis 6-7 km/s at offset distances greater than 10 km. An offset intraveltimes of the first refracted arrivals at -10 km and 11 km rangecan be seen in the OBDS and all the OBS's data. The gap results fromthe morphologically high relief along the shooting line in the reflec-tion experiment.

Trial-and-error forward modeling of the traveltimes was used todetermine the velocity structure. Theoretical traveltimes were calcu-lated by geometrical optics ray theory (Hirata and Shinjo, 1986;Cerveny and Psencfk, 1983). The initial 2-D model was constructedfrom 1-D models under individual OBS's. The sedimentary layer andthe acoustic basement were derived from the reflection data. The finalvelocity structure (Fig. 20) was determined so that the differencebetween calculated and observed traveltimes was less than 0.2 s;however, most calculated traveltimes fit with observed arrivals towithin 0.1 s. Our model can also explain the triplication seen at about10-km range on each seismometer.

Line EW

The OBDS, JRTl, JRT4, and JRT5 were used to determine thismodel (Figs. 21-24). The 4.5-km/s layer, which is laterally heteroge-

1081

M. SHINOHARA ET AL.

OBDS

in

Q

N

-

6 -

4 -

2 -

0 -

\

1

]

:j

i•

Hi

1H

I

i.

11

m ili

ffMSf

f1

Pj i

1111

f

11111111111 1lllllll §111111 III

raai

1|1

•II11

M

II

!1

yn

1

Si

•

;

t

<

-40 -30 -20 -10 0 10

Distance (km)

Figure 9. Record section of the OBDS for Line NS. See Figure 5 for detailed explanation.

20 30 40

JRT4

Q

8 -|

-

β -

-

4 -

2 -

0 -

N

π ri:

Iiiierr

p

iiI?• f n

PI

11II 11

gags

I1i1

s

Ii

1111

I

IPH •• 1• l_

-40 -30 -20 -10 0 10

Distance (km)

Figure 10. Record section of OBS JRT4 for Line NS. See Figure 5 for detailed explanation.

20 30 40

neous, is about 2 km thick along the profile. The 4.5-km/s layer is thinunder JRT1 east of the profile and thick under JRT5 west of the profile.The velocity of the layer below the 4.5-km/s layer is estimated to be6.2 km/s base on the tau-p method and the apparent velocities of thefirst arrivals on the OBS's. This model also predicts successfully thetriplication at 10-km distance on the OBDS and JRT4. Although thegradient of the 6.2-km/s layer could not be determined exactly, thegradient must be less than 0.2 s~1 because arrivals at offset distancesgreater than 25 km were observed on all seismometers. Because theOBDS observed arrivals with an apparent velocity of 6 km/s up to 40km, it is inferred that the 6.2-km/s layer extends downward to at least12-km depth from the sea surface (i.e., thickness is more than 6 km).

A layer with a P-wave velocity of 3.0 km/s overlies the 4.5-km/slayer around JRT5. The 3.0- km/s layer is required to explain thetraveltimes of refracted waves from the 4.5-km/s layer at distancesless than 5 km on JRT5.

Line NS

The OBDS, JRT4, and JRT8 were used to determine this model(Figs. 25-27). Layer thicknesses were forced to coincide with thoseof Line EW under the OBDS where two lines crossed. Because thenorth section of Line NS on the OBDS is not reversed, this part of themodel is not well constrained. The interface between the 4.5-km/s

1082


JRT8

LO

-40 -30 -20 -10 0 10

Distance (km)

Figure 11. Record section of OBS JRT8 for Line NS. See Figure 5 for detailed explanation.

layer and the 6.2-km/s layer is shallower at the both ends of thisprofile. A model with a 4.5-km/s layer with a thickness of about 2 kmand a 6.2-km/s layer can explain the triplication seen on all theseismometers. For the same reasons as in Line EW, the thickness ofthe 6.2-km/s layer is thought to be more than 6 km. Because the laterphase at offset distance greater than 45 km in the records of JRT8 isinterpreted as a crust-mantle reflection (PJP), the Moho is suggestedto be about 16 km deep.

DISCUSSION

The structure obtained in the northern Yamato Basin can becompared with structures in other areas in the Japan Sea (Fig. 28). Itis clear from our data that the northern Yamato Basin is different fromthe Japan Basin from crustal thickness alone. The crust in the JapanBasin is 8 km thick or less (Murauchi, 1972), whereas the data fromthe present study indicate no evidence of the oceanic crust. Althoughwe did not detect refracted arrivals from the Moho, the reflectionphase from the Moho (PmP) indicates that the depth should be greaterthan that in the Japan Basin.

The southern Yamato Basin was studied previously using modernreflection and refraction techniques that included profile lines shotwith an air gun and explosives to OBS's (Hirata et al., 1987, 1989;Katao, 1988; Chung et al., 1990). The crustal structure in the southernYamato Basin is neither typical oceanic nor continental. As the OBSair-gun profiling method used in the survey is almost identical to thatused in our experiment, it is worthwhile to compare directly bothrecords. Figure 29 shows a comparison between data from JRT8 inthe northern Yamato Basin and from YB7 (Fig. 1) in the southernYamato Basin. We see a close resemblance between the two recordsections over the first 25 km of range. Both crusts have a 1-km-thicksedimentary layer with a P-wave velocity from 2.0 to 3.0 km/s. Alayerwith a large velocity gradient underlies the sediments. The velocityin this layer increases from 3.0 to 5.2 km/s near Hole 794D, and from3.8 to 5.4 km/s beneath YB7 in the southern Yamato Basin (Fig. 30).The Ludwig et al. (1975) findings of a thick sediment and an uppercrust with a large velocity gradient in the northern Yamato Basin,using sonobuoy refraction data, also agree with our study.

The velocity structure from 7 to 12 km depth also is similar in boththe southern and northern Yamato Basin; the vertical gradient is0.1 s~1. We can not compare the lowermost crustal structure becauseno clear phases with reversed profiles are available in our data.

The total thickness of the crust can not be unambiguously deter-mined from our data. It is, however, likely that the northern YamatoBasin has slightly thinner crust (about 16 km below sea surface)because the reflected arrival from the Moho (PmP) at the JRT8 appearsat a shorter offset range (40 km) than was observed on YB7 in thesouthern Yamato Basin (45 km). These results are consistent withthose reported by Murauchi (1972) for profile P4 (Fig. 1) in thenorthern Yamato Basin. He found a 12-km-thick crust, (the Moho ata depth of 14 km below sea surface), indicating a thinner crust in thenorthern than in the southern Yamato Basin.

Thus, our observations and the previous studies indicate that thenorthern and southern Yamato Basin have a similar crustal structureas shown by the velocities and thicknesses of layers. The crustalstructure of both the basins are characterized as possessing a sedimen-tary layer about 1 km thick that is thinner than in the Japan Basin(Ludwig et al., 1975), and an uppermost igneous crust with a largevelocity gradient, previously not resolved.

Our detailed data analysis revealed that the velocity of the upper-most part of the 4.5-km/s layer increases rapidly from 2.0 to 4.5 km/sover 300 m of thickness. As we observed many sills and flows in thesampled core at Sites 794 and 797, the shallowmost crust of thebasement layer is interpreted to represent a sediment-sill complex,which is seismically imaged as a zone with large velocity gradientwith depth.

Recently it has been suggested that the thickened crust found inthe southern Yamato Basin formed because a large amount of igneousmaterial was intruded into it after an oceanic crust formed (Hirata etal., 1989). Another interpretation is that the thick crust (about 14 km)and the low velocity gradient of the 6.2-km/s layer may be consideredto indicate that an oceanic crust never formed through backarc evo-lution, and that the thinning of a continental crust during the initialstage of rifting is responsible for the obtained structure. In this case,igneous material intrusion was not so great. At this stage we cannotdiscriminate the two models from our results.

1083

M. SHINOHARA ET AL.

-40 -30

OBDS Horizontal Lina-NS

-20

Distance (km)

-40 -30

JRT4 Horizontal Linβ-NS

-20

Distance (km)

Figure 12. Comparison of the horizontal component record of the OBDS records to that of OBS (JRT4) just above the OBDS.Reduction velocity is 4.0 km/s. Band-pass filtering is 5-15 Hz. Each trace is normalized by its maximum amplitude.Converted 5-wave arrivals are seen in the records of the OBDS.

CONCLUSIONS

An ocean broadband downhole seismometer (OBDS) was in-stalled in Hole 794D in the northern Yamato Basin of the JapanSea as a part of Leg 128 in 1989. The digital signals of the OBDSwere continuously recorded on JOIDES Resolution for about 60hr, which we call the real-time experiment. A seismic refraction/re-flection experiment using air guns, the OBDS, and OBS's also was

conducted during the real-time experiment. Nine OBS's were de-ployed surrounding Hole 794D. Five OBS's were deployed every2800 m near Hole 794D, and the OBDS formed a 3-D sub array. Theprofiles consisted of two circles and two lines. Air guns were shotevery 60 s, i.e., 150 m. The signals of the air guns were also recordedby a single hydrophone streamer. The seismic structure was derivedfrom the data recorded by the OBDS and the four OBS's deployedevery 25 km.

1084


Velocity (km/s)

3 4 5

Velocity (km/s)

3 4 5

.o

Q.α>

Q

-Q

Q.α>

Q

Figure 13. One-dimensional P-wave velocity models beneath OBS JRT1obtained for each single-sided profile of Line EW by the tau-p method. Solidand dash lines correspond to the models derived from the records of the eastwardand westward shots, respectively. Depth is measured from the seafloor.

Velocity (km/s)

3 4 5

Figure 15. One-dimensional structure models of P-wave velocity beneath OBS

JRT5. See Figure 13 for detailed explanation. Layer with a P-wave velocity of

3.0^.5 km/s underlies the sedimentary layer.

Velocity (km/s)

i

Ha

Q.

Q

Figure 14. One-dimensional structure models of P-wave velocity beneath OBSJRT4. See Figure 13 for detailed explanation.

Figure 16. One-dimensional structure models of P-wave velocity beneath OBS

JRT4 obtained for each single-sided profile of Line NS. Solid and dash lines

correspond to the models derived from records of the northward and south-

ward shots, respectively.

1085

M. SHINOHARA ET AL.

Velocity (km/s)

3 4 5

w-Q

Q.

Q

Figure 17. One-dimensional structure models of P-wave velocity beneath OBSJRT8. See Figure 16 for detailed explanation.

One-dimensional seismic velocity structures beneath each OBSwere derived using the tau-p method with the direct tau-p transform.A sedimentary layer with a small velocity gradient exists just beneathseafloor. The sedimentary layer has a velocity of 1.7 km/s and is200-700 m thick. An acoustic basement with a P-wave velocity ofA.2-4.5 km/s underlies the sedimentary layer. A layer with a velocityof 3 km/s exists under the sedimentary layer only beneath JRT5.

Two-dimensional velocity structures along two linear profileswere derived by a forward modeling of the traveltimes using a 2-Dray-tracing method. The 4.5-km/s layer is 2-3 km thick. A layer withvelocity of 6.2 km/s underlies the 4.5-km/s layer. The velocity gradi-ent of the 6.2-km/s layer is less than 0.2 s ', and the thickness of the6.2-km/s layer appears to be more than 6 km.

It is clear that the northern Yamato Basin is different from theJapan Basin in crustal thickness. The Japan Basin has a typical oceaniccrustal thickness of 8 km. The upper crustal velocity structure of thenorthern Yamato Basin is similar to that of the southern Yamato Basin,where the crust is neither typical oceanic nor continental.

ACKNOWLEDGMENTS

We thank the scientists, staff, and crews of JOIDES Resolution,Tansei-maru, and Kaiko-maru 5 for their valuable contributions to theexperiment. The controlled-source seismic experiment was supported bythe following people, whom we thank: A. Nishizawa, Y. Kaiho, S. Abe,H. Shiobara, R. Hino, and S. Kuramoto. S. Ohmi provided the computerprogram for the data acquisition system. We thank L. S. L. Kong, whosereview helped to improve the manuscript, and K. Tamaki for discussion.The computation was done at the Information Processing Center of ChibaUniversity, the computer facilities of the University of Tokyo, and thecomputer room of Ocean Research Institute, the University of Tokyo.

REFERENCES

Abe, K., and Kanamori, H., 1970. Mantle structure beneath the Japan Searevealed by surface waves. Tokyo Daigaku Jishin Kenkyusho Iho,48:1011-1021.

Bee, M., and Bibee, L. D., 1989. A seismic refraction study of Cretaceousoceanic lithosphere in the northwest Pacific basin. Mar. Geophys. Res.,11:239-261.

Byrne, D. A., Harris, D., Duennebier, F. K., and Cessaro, R., 1987. The oceansub-bottom seismometer system installed in Deep Sea Drilling ProjectHole 581C, Leg 88: a technical review. In Duennebier, F. K., Stephen, R.,et al., Init. Repts. DSDP, 88: Washington (U.S. Govt. Printing Office),65-89.

Cerveny, V., and Psencik, I., 1983. Program SEIS83, Numerical Modeling ofSeismic Wave Fields in 2-D Laterally Varying Layered Structures by theRay Method: Praha (Charles Univ.).

Chung, T. W., Hirata, N., and Sato, R., 1990. Two-dimensional P- and 5-wavevelocity structure of the Yamato Basin, the southeastern Japan Sea, fromrefraction data collected by an ocean bottom seismographic array. J. Phys.Earth, 38:99-147.

Diebold, J. B., and Stoffa, P. L., 1981. The traveltime equation, tau-p mapping,and inversion of common midpoint data. Geophysics, 46:238-254.

Duennebier, F. K., Lienert, B., Cessaro, R., Anderson, P., and Mallick, S., 1987.Controlled-source seismic experiment at Hole 581C. In Duennebier, F. K.,Stephen, R., et al., Init. Repts. DSDP, 88: Washington (U.S. Govt. PrintingOffice), 105-125.

Evans, J. R., Suyehiro, K., and Sacks, I. S., 1978. Mantle structure beneath theJapan Sea—a re-examination. Geophys. Res. Lett, 5:487-490.

Hirata, N., Kinoshita, H., Suyehiro, K., Suyemasu, M., Matsuda, N., Ouchi,T., Katao, H., Koresawa, S., and Nagumo, S., 1987. Report on DELP 1985cruises in the Japan Sea Part II: seismic refraction experiment conductedin the Yamato Basin, southeast Japan Sea. Tokyo Daigaku JishinKenkyusho Iho, 62:347-365.

Hirata, N., and Shinjo, N., 1986. SEISOBS—modified version of SEIS83 forocean bottom seismograms. Jishin 2, 39:317-321. (in Japanese)

Hirata, N., Tokuyama, H., and Chung, T. W, 1989. An anomalously thicklayering of the crust of the Yamato Basin, southeastern Sea of Japan: thefinal stage of back-arc spreading. Tectonophysics, 165:303-314.

Ingle, J. C , Jr., Suyehiro, K., von Breymann, M. T., et al., 1990. Proc. ODP,Init. Repts., 128: College Station, TX (Ocean Drilling Program).

Isezaki, N., 1975. Possible spreading centers in the Japan Sea. Mar. Geophys.Res., 2:265-277.

Isezaki, N., Hata, K., and Uyeda, S., 1971. Magnetic survey of the Japan Sea(Part 1). Tokyo Daigaku Jishin Kenkyusho Iho, 49:77-83.

Isezaki, N., and Uyeda, S., 1973. Geomagnetic anomaly pattern of the JapanSea. Mar. Geophys. Res., 2:51-59.

Jacobson, R. S., Adair, R., and Orcutt, J., 1984. Preliminary seismic refractionresults using a borehole seismometer in Deep Sea Drilling Project HOLE395A. In Hyndman, R. D., Salisbury, M. H., et al., Init. Repts. DSDP, 78(Pt. 2): Washington (U.S. Govt. Printing Office), 783-792.

Kanazawa, T., 1986. Seven-component low-power acoustic release oceanbottom seismograph. Annu. Meet. Seismol. Soc. Jpn. Abstr., 2. (Abstract,in Japanese)

Katao, H., 1988. Seismic structure and formation of the Yamato Basin. TokyoDaigaku Jishin Kenkyusho Iho, 63:51-86.

Little, S. A., and Stephen, R. A., 1985. Costa Rica Rift borehole seismicexperiment, Deep Sea Drilling Project Hole 504B, Leg 92. In Anderson,R. N., Honnorez, J., etal., Init. Repts. DSDP, 83: Washington (U.S. Govt.Printing Office), 517-528.

Ludwig, W. L., Murauchi, S., and Houtz, R. E., 1975. Sediments and structureof the Japan Sea. Geol. Soc. Am. Bull., 86:651-664.

Matsuda, N., Fujii, T., and Kinoshita, H., 1986. Pop-up ocean bottom seis-mometer with hydrophone system. Annu. Meet. Seismol. Soc. Jpn. Abstr.,2:241. (Abstract, in Japanese)

Murauchi, S., 1972. Seismic crustal structure of the Japan Sea by controlled-source experiment. Kagaku, 42:367-375. (in Japanese)

Neidell, N. S., and Taner, M. T., 1971. Semblance and other coherencymeasures for multichannel data. Geophysics, 36:483^97.

Okada, H., Moriya, T., Masuda, T., Hasegawa, T., Asano, S., Kasahara, K.,Ikami, A., Aoki, H., Sasaki, Y, Furukawa, N., and Matsumura, K., 1978.

1086


Velocity anisotropy in the Sea of Japan as revealed by big explosions. J.Phys. Earth, 26, Suppl.:491-502.

Otofuji, Y., and Matsuda, T., 1984. Timing of rotational motion of SouthwestJapan inferred from paleomagnetism. Earth Planet. Sci. Lett., 70:373-382.

, 1987. Amount of clockwise rotation of Southwest Japan—fanshaped opening of the southwestern part of the Japan Sea. Earth Planet.Sci. Lett., 85:289-301.

Otofuji, Y., Matsuda, T., and Nohda, S., 1985. Paleomagnetic evidence forMiocene counter-clockwise rotation of Northeast Japan—rifting processof the Japan Arc. Earth Planet. Sci. Lett., 75:265-277.

Seama, N., and Isezaki, N., 1990. Sea-floor magnetization in the eastern part ofthe Japan Basin and its tectonic implications. Tectonophysics, 181:285-297.

Shearer, P., and Orcutt, J., 1985. Anisotropy in the oceanic lithosphere—theoryand observations from the Ngendei seismic refraction experiment in thesouth-west Pacific. Geophys. J. R. Astron. Soc, 80:493-526.

Shearer, P. M., Adair, R. G., Orcutt, J. A., and Jordan, T. H., 1986a. Simulta-neous borehole and ocean bottom seismometer recordings of earthquakesand explosions: results from the 1983 NGENDEI experiment at Deep SeaDrilling Project Hole 595B. In Duennebier, F. K., Stephen, R., Gettrust, J.F., et al., Init. Repts. DSDP, 91: Washington (U.S. Govt. Printing Office),377-383.

Shearer, P. M., Orcutt, J. A., Jordan, T. H., Whitmarsh, R. B., Kim, 1.1., Adair,R. G., and Burnett, M. S., 1986b. The NGENDEI seismic refractionexperiment at deep sea drilling project Hole 595B—ocean bottom seis-mometer data and evidence for crustal and upper mantle anisotropy, InDuennebier, F. K., Stephen, R., Gettrust, J. F., et al., Init. Repts. DSDP, 91:Washington (U.S. Govt. Printing Office), 385-399.

Shinohara, M., Amishiki, T., Hino, R., Hirata, N., and Kinoshita, H., 1989.Digital data acquisition system for single channel reflection experimentusing a micro computer. Annu. Meet. Seismol. Soc. Jpn. Abstr., 2:259.(Abstract, in Japanese)

Stephen, R. A., 1985. Seismic anisotropy in the upper oceanic crust. J.Geophys. Res., 90:11383-11396.

, 1988. Lateral heterogeneity in the upper oceanic crust at Deep SeaDrilling Project site 504. J. Geophys. Res., 93:6571-6584.

Stephen, R. A., and Harding, A. J., 1983. Travel time analysis of boreholeseismic data. J. Geophys. Res., 88:8289-8298.

Stephen, R. A., Johnson, S., and Lewis, B., 1983. The oblique seismicexperiment on Deep Sea Drilling Project Leg 65. In Lewis, B.T.R.,Robinson, P., et al., Init. Repts. DSDP, 65: Washington (U.S. Govt. PrintingOffice), 319-327.

Stoffa, P. L., Buhl, P., Diebold, J. B., and Wenzel, E, 1981. Direct mapping ofseismic data to the domain of intercept time and ray parameter—a plane-wave decomposition. Geophysics, 46:255-267.

Swift, S. A., and Stephen, R. A., 1989. Lateral heterogeneity in the seismicstructure of upper oceanic crust, Western North Atlantic. J. Geophys. Res.,94:9303-9322.

Swift, S. A., Stephen, R. A., and Hoskins, H., 1988. Structure of upper oceaniccrust from an oblique seismic experiment at Hole 418A, Western NorthAtlantic. In Salisbury, M. H., Scott, J. H., et al., Proc. ODP, Sci. Results,102: College Station, TX (Ocean Drilling Program), 97-124.

Tamaki, K., 1986. Age estimation of the Japan Sea on the basis of stratigraphy,basement depth, and heat flow data. J. Geomagn. Geoelectr., 38:427-446.

Urabe, T., and Hirata, N., 1984. A playback system for long-term analog taperecordings of ocean bottom seismographs. Jishin 2, 37:633-645. (inJapanese)

Whitmarsh, R. B., Orcutt, J. A., Jordan, T. H., Adair, R. G., and Shearer, P. M.,1986. Velocity bounds on the seismic structure of Mesozoic crust and uppermantle in the southwest Pacific basin from downhole observations at DeepSea Drilling Project Hole 595B. In Duennebier, F. K., Stephen, R.,Gettrust, J. E, et al., Init. Repts. DSDP, 91: Washington (U.S. Govt.Printing Office), 437^144.

Yamada, T., 1980. Spatial distribution of microearthquakes and crust in thesubduction area revealed by ocean seismographic observations [Ph.D.dissert.]. Univ. Tokyo, (in Japanese)

Yamano, M., Uyeda, S., Uyeshima, M., Kinoshita, M., Nagihara, S., Boh, R.,and Fujisawa, H., 1987. Report on DELP 1985 cruises in the Japan Sea.Part V: heat flow measurements. Tokyo Daigaku Jishin, Kenkyusho Iho,62:417^132.

Yasui, M., Kishii, T., Watanabe, T., and Uyeda, S., 1968. Heat flow in the Seaof Japan. In The Crust and Upper Mantle of the Pacific Area. Am.Geophys. Union Monogr., 12:3-16.

Yoshii, T., 1972. Terrestrial heat flow and features of the upper mantle beneaththe Pacific and Sea of Japan. J. Phys. Earth, 20:271-285.

Date of initial receipt: 26 March 1991Date of acceptance: 1 September 1991Ms 127/128B-230

1087

M. SHINOHARA ET AL.

W JRT4

•SH•

;

nil

i

-10 0 10 Velocity (km/s)

Distance (km)Figure 18. Comparison of the reflection records with the tau-p method results for JRT4 Line EW. The one-dimensional structure model of Figure 14 has beenconverted to two-way traveltime measured from the seafloor. Solid and dash lines correspond to the models derived from records of the eastward and westwardshots, respectively. Position of OBS is marked. P-wave velocity rapidly increases at the acoustic basement. P-wave velocities in the sedimentary layer anduppermost part of acoustic basement are 1.7 and 4.5 km/s, respectively.

1088


W

-30 •15 0 15

Distance (km)

30 45

Figure 19. Examples of traveltimes of first arrivals on the OBDS and OBS'sfor Line EW. Origin of abscissa is at the OBDS position. Solid triangles showthe positions of each OBS. A phase with an apparent velocity of about 6.0-7.0km/s is seen in all records.

Q.

Q 11

14

17

6.8

Moho ?(km/s)

-45 -35 -25 -15 -5 5 15 25 35 45

Distance (km)

BN

14"

17

JRT4

(km/s)

Moho?

-35 -25 - 1 5 - 5 5 15

Distance (km)

25 35

Figure 20. P-wave velocity structure derived for Line EW (A) and Line NS (B)in the northern Yamato Basin. Numerals are P-wave velocities in km/s. Trianglesshow the positions of the OBDS and OBS's. The OBDS is located at theintersection of Line EW and Line NS. Vertical exaggeration is about 1.5×.

1089

M. SHINOHARA ET AL.

W

-45 -35 -25 -15 -5 5

Distance (km)

15 25 45

Figure 21. Results of ray tracing using the Line EW model to the OBDS. Distance (D) along the profile Line EW is measured from the OBDS. The OBDS is atD = 0.0 km. A: Ray path diagrams. B: Observed traveltimes (solid squares) with arrival time reading errors and calculated traveltimes for the model (other symbols).Reading errors of later arrivals are more than 0.5 s and are not shown. Open squares indicate traveltime of water waves. Crosses, triangles, and circles denotetraveltimes of reflection and refraction waves from basement, 4.5-km layer, and 6.2-km layer, respectively.

1090


B

W

3-

2-

1 -

π Direct* Basement

4.5km/s layerO 6.2km/s layer• Observation

-45 -35 -25

Figure 21 (continued).

o o

-15 -5 5 15 25

Distance (km)

35 45

1091

M. SHINOHARA ET AL.

W

-45 -35 -25 -15 -5 5

Distance (km)

15 25 35

Figure 22. Results of ray tracing using the Line EW model to OBS JRT4 located at D = 0.0 km. See Figure 21 for detailed explanation.

1092

YAMATO BASIN CRUST AL STRUCTURE

W

3"

2-

1-

π Direct* Basement

4.5km/s layer6.2km/s layerObservation

o o

-45 -35 -25 -15 -5 5 15 25 35 45

Distance (km)


1093

M. SHINOHARA ET AL.

W

-35 -25 -15-45

Figure 23. Results of ray tracing using the Line EW model to OBS JRTl located at D = 24.3 km. See Figure 21 for detailed explanation. Pluses indicate traveltimeof Pn phase.

1094


i ß2"

1-

W

D Direct× Basement

4.5km/s layero 6.2km/s layer+ Moho• Observation

O

1 Δ

\J

-45 -35 -25 -15 -5 5

Distance (km)

15 25 35 45


1095

M. SHINOHARA ET AL.

w

Q.ΦQ

-45 -35 -25 -15 -5 5

Distance (km)

Figure 24. Results of ray tracing using the Line EW model to OBS JRT5 located at D = -23.7 km. See Figure 21 for detailed explanation.

1096

YAMATO BASIN CRUST AL STRUCTURE

w

3-

1-

α Direct* Basement


Φ

O O

o Φ 0o δ

-45


-35 -25 -15 -5 5

Distance (km)

15 25 35 45

1097

M. SHINOHARA ET AL.

N

Iα.Q

-35 -25

Figure 25. Results of ray tracing using the Line NS to model to the OBDS located at D = O.O km. See Figure 21 for detailed

explanation.

1098


N

3-

Qi

1-

-35 -25 -15


-5 5

Distance (km)

α Direct× Basement

4.5km/s layero 6.2km/s layer• Observation

15 25 35

1099

M. SHINOHARA ET AL.

N

-35 -25 -15 -5 5 15 25 35

Distance (km)

Figure 26. Results of ray tracing using the Line NS model to OBS JRT4 located at D = -0.4 km. See Figure 21 for detailed explanation.

1100


N

3"

Qi

1-

-35 -25


-15 -5 5

Distance (km)

D Direct* Basement

4.5km/s layero 6.2km/s layer• Observation

15 25 35

1101

M. SHINOHARA ET AL.

N

-25 -15 -5 5 15 25 35

Distance (km)

Figure 27. Results of ray tracing using the Line NS model to OBS JRT8 located at D = 25.0 km. See Figure 21 for detailed explanation.

1102


N

3-

mQ

11

\J

-35


-25 -15

D Direct× Basement


-5 5

Distance (km)

15 25 35

1103

M. SHINOHARA ET AL.

10

14

18

(km)

SouthernYamato Basin

Hirata et al.(1989)

|6.7 - 7.11

7.1 - 7.2

7.2 - 7.4

7.8(km/s)

YB7

4.4 - 4.6

NorthernYamato Basin

This studySite794D

Murauchi(1972)

Moho

HUH 9 iiiiiiin 9 inn

(km/s)

OBDSJRT4

.0

4.2

6.2

- [6-7] -

7.7

(km/s)

P4

Japan Basin

Murauchi(1972)

1.5

* 2.13.0

4.8

6.7

8.1

10

14

(km/s)-"18

P1 (km)

Figure 28. Comparison of the seismic structure of the northern Yamato Basin to structures in other areas of the Japan Sea. YB7 is an OBS station in the 1985OBS's air-gun and explosion experiment (Hirata et al., 1987). PI and P4 (Murauchi, 1972) show the results of the two-ship seismic experiments (locations ofYB7, PI, and P4 in Fig. 1). The upper crustal structure of the northern Yamato Basin is similar to that of the southern Yamato Basin. A thick sedimentary layerand an upper crust with a large velocity gradient characterize the seismic structure of the Yamato Basin.

1104


sw NE

40 20Distance (km)

Southern Yamato Basin

40 20Distance (km)

Northern Yamato Basin

YB-7

0JRT8

Figure 29. Comparison of seismic data (vertical component, high-gain channel) from JRT8 in the northern Yamato Basin with that from YB7 inthe southern Yamato Basin. Reduction velocity is 6.0 km/s. Band-pass filtering is 5-20 Hz. Records for the southwestward shots of YB7 andnorthward shots of JRT8 are plotted. A close resemblance between the two record sections is seen at ranges less than 25 km.

1105

M. SHINOHARA ET AL.

Q.Q)Q

u

5

10

15

on

x ^\

-

-

This study

Hirataetal.

i i •

\λ —

(1989)

i

•

\

\ - \

\ \

\ \

\ \

\ \

\ \

x \

? i

i

V

0 2 4 6 8

Velocity (km/s)

Figure 30. Comparison of P-wave velocity-depth functions for the northernYamato Basin (solid line) and the southern Yamato Basin (dash line). Bothcrusts have a 1 -km-thick sedimentary layer that overlies a layer with a largevelocity gradient.

1106

Documents

Ocean Drilling Program: Science Operator - 69. …...In September 1989, an ocean broadband digital downhole seismometer (OBDS) was installed at Hole 794D in the northern Yamato Basin